基于改进粒子群算法的神经网络PID控制研究

BP神经网络PID控制是利用BP神经网络的自学习和逼近任意非线性函数功能,对PID控制器的三个参数进行在线整定,但网络初始权值的选取困难.采用改进的PSO算法优化BP神经网络的初始权值,并对基于PAO算法的BP神经网络PID控制进行仿真实验.仿真结果表明,PSO算法使得网络初始权值的选取比较快速,系统的性能有所提高.

Research on Neural Network PID Control Based on Improved PSO

The neural network has the capability of self-studying and approximating to any nonlinear functions.BP neural network PID can adjust three PID parameters on line.But the selection of initial weight coefficients of the BP neural network is difficult.In this paper,the improved PSO algorithm is adopted to optimize initial weight coefficients of the BP neural network,and simulation test on neural network PID control based on PSO algorithm is carried out.The simulation results show the selection of initial weight coefficients of the BP neural network is fast relatively,and the system has a better performance.